Team Workflow aggiornato #37
@@ -41,7 +41,7 @@ api:
|
|||||||
agents:
|
agents:
|
||||||
|
|
|||||||
strategy: Conservative
|
strategy: Conservative
|
||||||
team_model: qwen3:1.7b
|
team_model: qwen3:1.7b
|
||||||
team_leader_model: qwen3:4b
|
team_leader_model: qwen3:8b
|
||||||
query_analyzer_model: qwen3:4b
|
query_analyzer_model: qwen3:4b
|
||||||
report_generation_model: qwen3:4b
|
report_generation_model: qwen3:8b
|
||||||
predictor_model: qwen3:4b
|
predictor_model: qwen3:4b
|
||||||
|
|||||||
@@ -44,7 +44,7 @@ class PipelineInputs:
|
|||||||
"""
|
"""
|
||||||
Sceglie il modello LLM da usare per il Team Leader.
|
Sceglie il modello LLM da usare per il Team Leader.
|
||||||
"""
|
"""
|
||||||
self.leader_model = self.configs.models.all_models[index]
|
self.team_leader_model = self.configs.models.all_models[index]
|
||||||
|
|
||||||
def choose_team(self, index: int):
|
def choose_team(self, index: int):
|
||||||
|
Variable assignment uses Variable assignment uses `team_leader_model` instead of `leader_model` as mentioned in the method's docstring. The docstring should be updated or the variable name should be consistent.
The The `choose_team_leader` method assigns to `self.team_leader_model` but the initialization in `__init__` uses a different assignment pattern via `get_model_by_name()`. This inconsistency means the model assignment won't have the same validation and could break if `index` is out of bounds. Use `self.team_leader_model = self.configs.models.all_models[index]` with proper validation or use the same pattern as `__init__`.
```suggestion
model_list = self.configs.models.all_models
if 0 <= index < len(model_list):
model_name = model_list[index].name
self.team_leader_model = self.configs.get_model_by_name(model_name)
else:
raise IndexError(f"Model index {index} out of range for team leader selection.")
```
|
|||||||
"""
|
"""
|
||||||
|
|||||||
Reference in New Issue
Block a user
The
predictor_modelconfiguration field appears to be unused after the refactoring that replaced it withquery_analyzer_modelandreport_generation_model. This configuration should be removed to avoid confusion.